Kimi K2 is a state-of-the-art mixture-of-experts (MoE) language model with 32 billion activated parameters and 1 trillion total parameters. Trained with the MuonClip optimizer, it achieves exceptional performance across frontier knowledge, reasoning, and coding tasks while being meticulously optimized for agentic capabilities. The instruct variant is post-trained for drop-in, general-purpose chat and agentic experiences without long thinking.
发布日期2025年7月11日
参数规模1.0T
上下文长度131K
许可证MIT
知识截止—
Benchmarks
评测成绩
| 评测基准 | 类别 | 分数 | 来源 |
|---|---|---|---|
| MATH-500 | math reasoning | 97.4 | 来源 |
| GSM8k | math reasoning | 97.3 | 来源 |
| CBNSL | math reasoning | 95.6 | 来源 |
| HumanEval | reasoning code | 93.3 | 来源 |
| MMLU-Redux | language reasoning math general | 92.7 | 来源 |
| IFEval | general | 89.8 | 来源 |
| AutoLogi | reasoning | 89.5 | 来源 |
| MMLU | general reasoning language math | 89.5 | 来源 |
| ZebraLogic | reasoning | 89.0 | 来源 |
| MultiPL-E | general language | 85.7 | 来源 |
| MMLU-Pro | language reasoning math general | 81.1 | 来源 |
| HumanEval-ER | reasoning | 81.1 | 来源 |
| CSimpleQA | general language | 78.4 | 来源 |
| AceBench | general reasoning | 76.5 | 来源 |
| LiveBench | math reasoning general | 76.4 | 来源 |
| MuSR | reasoning | 76.4 | 来源 |
| GPQA | reasoning general | 75.1 | 来源 |
| CNMO 2024 | math | 74.3 | 来源 |
| SWE-bench Verified (Multiple Attempts) | reasoning | 71.6 | 来源 |
| Tau2 retail | communication reasoning | 70.6 | 来源 |
| AIME 2024 | math reasoning | 69.6 | 来源 |
| Tau2 telecom | communication reasoning | 65.8 | 来源 |
| SWE-bench Verified (Agentic Coding) | reasoning code | 65.8 | 来源 |
| PolyMath-en | math reasoning | 65.1 | 来源 |
| Aider-Polyglot | general code | 60.0 | 来源 |
| SuperGPQA | reasoning general math legal healthcare finance chemistry economics physics | 57.2 | 来源 |
| Tau2 airline | reasoning communication | 56.5 | 来源 |
| MultiChallenge | communication reasoning | 54.1 | 来源 |
| LiveCodeBench v6 | reasoning general | 53.7 | 来源 |
| SWE-bench Verified (Agentless) | general reasoning | 51.8 | 来源 |
| AIME 2025 | math reasoning | 49.5 | 来源 |
| SWE-bench Multilingual | reasoning code | 47.3 | 来源 |
| HMMT 2025 | math | 38.8 | 来源 |
| SimpleQA | general reasoning | 31.0 | 来源 |
| Terminal-bench | reasoning code | 30.0 | 来源 |
| OJBench | reasoning | 27.1 | 来源 |
| Terminus | reasoning code | 25.0 | 来源 |
| Humanity's Last Exam | general | 4.7 | 来源 |
Pricing
API 价格对比
| 服务商 | 输入价 | 输出价 | 上下文 | 吞吐(tok/s) | 延迟(s) | 函数调用 | 代码执行 | 联网搜索 |
|---|---|---|---|---|---|---|---|---|
| Cortecs | $0.55 | $2.65 | 131K | — | — | ✓ | ✗ | ✗ |
价格单位：美元/百万 token，数据来自社区整理，仅供参考。